The Bayesian estimation of the parameters and reliability function for Wiener process are obtained based on the generalized entropy loss function by using both non-informative and conjugate prior distribution,and it is compared with the Bayesian estimation under the square loss function and maximum likelihood estimation. The simulation results show that: The Bayesian estimation under the generalized entropy loss function has the smallest mean square error and the highest precision, and the expression of the estimation is flexible, which can effectively describe the situation where the risks are different due to over-estimation and under-estimation.